DESCRIPTION (Verbatim from the Applicant's Abstract): Diagnostic imaging of cancer has been identified by the NCI as a priority program principally because of the commonly held belief that if cancer can be more accurately located in the body, then patient management and possible outcome will improve. This application will develop a new imaging modality, namely a bispecific antibody (BsMAb) approach that will be suitable for gamma and PET imaging. Since many positron-emitters have a shorter half-life than the gamma-emitters, the method must be able to provide images within a short time after the isotope injection. Preliminary results with the BsMAb approach used in this application indicate imaging will be possible with I to 3 hours. Thus, this methodology should be suitable for either a gamma- or a positron-emitting agent. With superior tumor/non/tumor ratios, this method could improve the sensitivity and specificity of diagnostic cancer imaging by gamma imaging or PET (ImmunoPET). This proposal includes preclinical studies that will continue to build on the methodology described in this application. F(ab')2 x Fab' and Fab' x Fab' BsMAb constructs will be explored to deterrnine the optimal agent for targeting. Clinical studies are proposed by the middle of the second year. The clinical studies will use a humanized anti-carcinoembryonic antigen (CEA) antibody (hMN-14) that is chemically coupled to a murine antiDTPA-(In) antibody to form a BsMAb fragment (i.e., Fab' x Fab' or F(ab')2 x Fab'). Preclinical studies have already shown this BsMAb targeting method together with a technetium-binding di-DTPA peptide can localize colon cancer xenografts within 1 hour of the peptide's injection. Clinical studies will first examine parameters designed to optimize this approach using 99mTc-peptide (gamma-imaging method). Once the methodology has been optimized clinically, future studies (not included in this application) will begin to focus on a larger cohort of patients to test the targeting efficacy using the 99mTc-labeled peptide, but we would also propose to initiate development of a PET imaging system using this same methodology and possibly 94TnTc.